~/ services/ venture studio
service · 03VENTURE STUDIOequity · co-found

1–4 ventures
per year.

We don't run an accelerator. We co-found one to four AI ventures a year with domain operators we trust, write a real first cheque, and pour the lab's full research stack into the foundation. Six in four years. Every one still operating.

The thesis

Most venture builders run a high-cadence factory: 6 — 8 companies a year, thin support, optimise for the law of large numbers. We do the opposite. One to four companies a year, deep co-founding, sized to the lab's actual capacity. A studio that ships fewer companies but treats each one like the only one.

Our pattern: a domain operator with a thesis we believe, paired with a lab pod that builds the technical core. The operator runs the company. The lab is the technical co-founder for the first 18 months and a board observer thereafter.

What we bring to day one

// 01

First cheque

€250k — €2m on standard SAFE. Wired the day the company incorporates.

// 02

Founding squad

3 — 5 lab engineers seconded for 18 months. Some convert to permanent hires.

// 03

Research stack

Full access to the lab's code, datasets, eval harnesses, and internal models.

// 04

Operator network

Warm intros to design partners, follow-on investors, and adjacent-domain operators.

// 05

Back-office

Legal, finance, hiring, ops — handled by the lab for the first 12 months.

// 06

Patient board seat

One board observer from the lab. We don't churn boards. We don't push exits.

What we look for in an operator

The studio cohort skews one direction: people who've worked the domain for a decade and have spent the last year quietly going insane about the gap they want to close. We're picky about three things.

  1. Domain depth. Not a research postdoc with a paper. Someone who's seen the problem fail in production, multiple times, at scale.
  2. Operator instincts. You've shipped something. To customers. With a P&L. The default is to go fast and iterate, not to wait for the model.
  3. Comfort with the lab. You want a technical co-founder, not a contractor. You're willing to share equity for serious leverage.

From conversation to incorporation

// stage 01 · 4 weeks

Conversations
  • Three working sessions
  • Thesis stress-test
  • Mutual diligence

// stage 02 · 8 weeks

Pre-build
  • Joint research sprint
  • Cap table modelling
  • Term-sheet drafted

// stage 03 · day 0

Incorporation
  • Co. formed
  • SAFE wired
  • Squad seconded

Terms — without the dance

  • VehicleYC post-money SAFE, no MFN games
  • Check size€250k — €2m, sized to the 18-month plan
  • Equity15 — 25% post-money on conversion
  • Founder splitoperator-led; we are minority co-founders
  • Vesting4 years, 1-year cliff, both sides
  • Board2 founders + 1 lab observer; no investor seat at this stage
  • Pro-ratafull pro-rata in the next two rounds
  • Exit termsnone. we are patient.

The cohort

Six companies, four years, six different domains — marine robotics, soilless agriculture, smart power, industrial kitchens, cheminformatics, acoustic AI for defense. Read the case studies and decide whether this is the company you want yours to keep.

6/6still operating
$32m+capital deployed across cohort
4 yrsvintage range, 2022–2025

read the cohort →

have a thesis
worth co-founding?

Send a one-pager — your domain, the gap, why now. Three working sessions and we'll know whether to build it together.